On-Center and Off-Center Cell Competition Generates Oriented Receptive Fields from Non-Oriented Stimuli in Kohonen’s Self-Organizing Map

  • Maximilian Riesenhuber
  • Hans-Ulrich Bauer
  • Theo Geisel


The self-organization of sensotopic maps, in particular of visual maps, continues to be an area of great interest in computational neuroscience. In order to distinguish between the different map formation models and between the specific self-organization mechanisms they assume, their behavior with regard to an as large number of physiological, anatomical or theoretical constraints as possible has to be investigated. An interesting case in point is the development of oriented receptive fields from stimulus distributions or stimuli with rotational symmetry, i.e., without orientation. This physiologically quite plausible symmetry breaking phenomenon has been observed in several models for the self-organization of receptive fields [1] orientation maps [2,3] which assumed a competition of On-center and Off-center cells,with rotational symmetry of the stimulus autocorrelation function. Whereas these models are characterized by a linear kernel operating on the afferent activity distribution and mediating the lateral interaction, a major competing model, Kohonen’s Self-Organizing Map (SOM, [4]) employs a strongly non-linear lateral interaction function. This nonlinearity is presumably responsible for the successful reproduction of several properties of visual maps in respective SOM-models, like the widening of ocular dominance bands as a consequence of decreased stimulus correlation [5,6], or the preferred angle of intersection between iso-orientation and iso-ocularity bands [7, 8]. However, the investigation of SOM-models with regard to the formation of oriented receptive fields upon stimulation with non-oriented On-center and Off-center stimuli turned out to be difficult, not the least because these models are numerically expensive to simulate.


Receptive Field Rotational Symmetry Computational Neuroscience Layer Symmetry Stimulus Distribution 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    R. Linsker, Proc. Nat. Acad. Sci. USA 83, 8390 (1986).PubMedCrossRefGoogle Scholar
  2. 2.
    K. Miller, J. Neurose. 14, 409 (1994).Google Scholar
  3. 3.
    M. Miyashita, S. Tanaka, NeuroRep. 3, 69 (1992).CrossRefGoogle Scholar
  4. 4.
    T. Kohonen, Senf Organizing Maps, Springer, Berlin (1995).CrossRefGoogle Scholar
  5. 5.
    G. J. Goodhill, Biol. Cyb. 69, 109 (1993).CrossRefGoogle Scholar
  6. 6.
    H.-U. Bauer, M. Riesenhuber, T. Geisel. Phys. Rev. E54, 2807–2810 (1996)Google Scholar
  7. 7.
    E. Erwin, K. Obermayer, K. Schulten, Neur. Comp. 7, 425 (1995).CrossRefGoogle Scholar
  8. 8.
    F. Hoffsümmer, F. Wolf, T. Geisel, S. Löwel, K. Schmidt, Proc. CNS 1995 Monterey, eds. J. Bower, Kluwer, Boston (1996), p. 197Google Scholar

Copyright information

© Springer Science+Business Media New York 1997

Authors and Affiliations

  • Maximilian Riesenhuber
    • 1
  • Hans-Ulrich Bauer
    • 1
  • Theo Geisel
    • 1
  1. 1.Institut für Theoretische Physik SFB Nichtlineare DynamikUniversität FrankfurtFrankfurt/MainGermany

Personalised recommendations